1 /* 2 * Copyright (c) 2006 Oracle. All rights reserved. 3 * 4 * This software is available to you under a choice of one of two 5 * licenses. You may choose to be licensed under the terms of the GNU 6 * General Public License (GPL) Version 2, available from the file 7 * COPYING in the main directory of this source tree, or the 8 * OpenIB.org BSD license below: 9 * 10 * Redistribution and use in source and binary forms, with or 11 * without modification, are permitted provided that the following 12 * conditions are met: 13 * 14 * - Redistributions of source code must retain the above 15 * copyright notice, this list of conditions and the following 16 * disclaimer. 17 * 18 * - Redistributions in binary form must reproduce the above 19 * copyright notice, this list of conditions and the following 20 * disclaimer in the documentation and/or other materials 21 * provided with the distribution. 22 * 23 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, 24 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF 25 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND 26 * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS 27 * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN 28 * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN 29 * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE 30 * SOFTWARE. 31 * 32 */ 33 #include <linux/kernel.h> 34 #include <linux/in.h> 35 #include <linux/if.h> 36 #include <linux/netdevice.h> 37 #include <linux/inetdevice.h> 38 #include <linux/if_arp.h> 39 #include <linux/delay.h> 40 #include <linux/slab.h> 41 42 #include "rds.h" 43 #include "ib.h" 44 45 unsigned int fmr_pool_size = RDS_FMR_POOL_SIZE; 46 unsigned int fmr_message_size = RDS_FMR_SIZE + 1; /* +1 allows for unaligned MRs */ 47 unsigned int rds_ib_retry_count = RDS_IB_DEFAULT_RETRY_COUNT; 48 49 module_param(fmr_pool_size, int, 0444); 50 MODULE_PARM_DESC(fmr_pool_size, " Max number of fmr per HCA"); 51 module_param(fmr_message_size, int, 0444); 52 MODULE_PARM_DESC(fmr_message_size, " Max size of a RDMA transfer"); 53 module_param(rds_ib_retry_count, int, 0444); 54 MODULE_PARM_DESC(rds_ib_retry_count, " Number of hw retries before reporting an error"); 55 56 struct list_head rds_ib_devices; 57 58 /* NOTE: if also grabbing ibdev lock, grab this first */ 59 DEFINE_SPINLOCK(ib_nodev_conns_lock); 60 LIST_HEAD(ib_nodev_conns); 61 62 void rds_ib_add_one(struct ib_device *device) 63 { 64 struct rds_ib_device *rds_ibdev; 65 struct ib_device_attr *dev_attr; 66 67 /* Only handle IB (no iWARP) devices */ 68 if (device->node_type != RDMA_NODE_IB_CA) 69 return; 70 71 dev_attr = kmalloc(sizeof *dev_attr, GFP_KERNEL); 72 if (!dev_attr) 73 return; 74 75 if (ib_query_device(device, dev_attr)) { 76 rdsdebug("Query device failed for %s\n", device->name); 77 goto free_attr; 78 } 79 80 rds_ibdev = kmalloc(sizeof *rds_ibdev, GFP_KERNEL); 81 if (!rds_ibdev) 82 goto free_attr; 83 84 spin_lock_init(&rds_ibdev->spinlock); 85 86 rds_ibdev->max_wrs = dev_attr->max_qp_wr; 87 rds_ibdev->max_sge = min(dev_attr->max_sge, RDS_IB_MAX_SGE); 88 89 rds_ibdev->fmr_max_remaps = dev_attr->max_map_per_fmr?: 32; 90 rds_ibdev->max_fmrs = dev_attr->max_fmr ? 91 min_t(unsigned int, dev_attr->max_fmr, fmr_pool_size) : 92 fmr_pool_size; 93 94 rds_ibdev->max_initiator_depth = dev_attr->max_qp_init_rd_atom; 95 rds_ibdev->max_responder_resources = dev_attr->max_qp_rd_atom; 96 97 rds_ibdev->dev = device; 98 rds_ibdev->pd = ib_alloc_pd(device); 99 if (IS_ERR(rds_ibdev->pd)) 100 goto free_dev; 101 102 rds_ibdev->mr = ib_get_dma_mr(rds_ibdev->pd, 103 IB_ACCESS_LOCAL_WRITE); 104 if (IS_ERR(rds_ibdev->mr)) 105 goto err_pd; 106 107 rds_ibdev->mr_pool = rds_ib_create_mr_pool(rds_ibdev); 108 if (IS_ERR(rds_ibdev->mr_pool)) { 109 rds_ibdev->mr_pool = NULL; 110 goto err_mr; 111 } 112 113 INIT_LIST_HEAD(&rds_ibdev->ipaddr_list); 114 INIT_LIST_HEAD(&rds_ibdev->conn_list); 115 list_add_tail(&rds_ibdev->list, &rds_ib_devices); 116 117 ib_set_client_data(device, &rds_ib_client, rds_ibdev); 118 119 goto free_attr; 120 121 err_mr: 122 ib_dereg_mr(rds_ibdev->mr); 123 err_pd: 124 ib_dealloc_pd(rds_ibdev->pd); 125 free_dev: 126 kfree(rds_ibdev); 127 free_attr: 128 kfree(dev_attr); 129 } 130 131 void rds_ib_remove_one(struct ib_device *device) 132 { 133 struct rds_ib_device *rds_ibdev; 134 struct rds_ib_ipaddr *i_ipaddr, *i_next; 135 136 rds_ibdev = ib_get_client_data(device, &rds_ib_client); 137 if (!rds_ibdev) 138 return; 139 140 list_for_each_entry_safe(i_ipaddr, i_next, &rds_ibdev->ipaddr_list, list) { 141 list_del(&i_ipaddr->list); 142 kfree(i_ipaddr); 143 } 144 145 rds_ib_destroy_conns(rds_ibdev); 146 147 if (rds_ibdev->mr_pool) 148 rds_ib_destroy_mr_pool(rds_ibdev->mr_pool); 149 150 ib_dereg_mr(rds_ibdev->mr); 151 152 while (ib_dealloc_pd(rds_ibdev->pd)) { 153 rdsdebug("Failed to dealloc pd %p\n", rds_ibdev->pd); 154 msleep(1); 155 } 156 157 list_del(&rds_ibdev->list); 158 kfree(rds_ibdev); 159 } 160 161 struct ib_client rds_ib_client = { 162 .name = "rds_ib", 163 .add = rds_ib_add_one, 164 .remove = rds_ib_remove_one 165 }; 166 167 static int rds_ib_conn_info_visitor(struct rds_connection *conn, 168 void *buffer) 169 { 170 struct rds_info_rdma_connection *iinfo = buffer; 171 struct rds_ib_connection *ic; 172 173 /* We will only ever look at IB transports */ 174 if (conn->c_trans != &rds_ib_transport) 175 return 0; 176 177 iinfo->src_addr = conn->c_laddr; 178 iinfo->dst_addr = conn->c_faddr; 179 180 memset(&iinfo->src_gid, 0, sizeof(iinfo->src_gid)); 181 memset(&iinfo->dst_gid, 0, sizeof(iinfo->dst_gid)); 182 if (rds_conn_state(conn) == RDS_CONN_UP) { 183 struct rds_ib_device *rds_ibdev; 184 struct rdma_dev_addr *dev_addr; 185 186 ic = conn->c_transport_data; 187 dev_addr = &ic->i_cm_id->route.addr.dev_addr; 188 189 rdma_addr_get_sgid(dev_addr, (union ib_gid *) &iinfo->src_gid); 190 rdma_addr_get_dgid(dev_addr, (union ib_gid *) &iinfo->dst_gid); 191 192 rds_ibdev = ib_get_client_data(ic->i_cm_id->device, &rds_ib_client); 193 iinfo->max_send_wr = ic->i_send_ring.w_nr; 194 iinfo->max_recv_wr = ic->i_recv_ring.w_nr; 195 iinfo->max_send_sge = rds_ibdev->max_sge; 196 rds_ib_get_mr_info(rds_ibdev, iinfo); 197 } 198 return 1; 199 } 200 201 static void rds_ib_ic_info(struct socket *sock, unsigned int len, 202 struct rds_info_iterator *iter, 203 struct rds_info_lengths *lens) 204 { 205 rds_for_each_conn_info(sock, len, iter, lens, 206 rds_ib_conn_info_visitor, 207 sizeof(struct rds_info_rdma_connection)); 208 } 209 210 211 /* 212 * Early RDS/IB was built to only bind to an address if there is an IPoIB 213 * device with that address set. 214 * 215 * If it were me, I'd advocate for something more flexible. Sending and 216 * receiving should be device-agnostic. Transports would try and maintain 217 * connections between peers who have messages queued. Userspace would be 218 * allowed to influence which paths have priority. We could call userspace 219 * asserting this policy "routing". 220 */ 221 static int rds_ib_laddr_check(__be32 addr) 222 { 223 int ret; 224 struct rdma_cm_id *cm_id; 225 struct sockaddr_in sin; 226 227 /* Create a CMA ID and try to bind it. This catches both 228 * IB and iWARP capable NICs. 229 */ 230 cm_id = rdma_create_id(NULL, NULL, RDMA_PS_TCP); 231 if (IS_ERR(cm_id)) 232 return PTR_ERR(cm_id); 233 234 memset(&sin, 0, sizeof(sin)); 235 sin.sin_family = AF_INET; 236 sin.sin_addr.s_addr = addr; 237 238 /* rdma_bind_addr will only succeed for IB & iWARP devices */ 239 ret = rdma_bind_addr(cm_id, (struct sockaddr *)&sin); 240 /* due to this, we will claim to support iWARP devices unless we 241 check node_type. */ 242 if (ret || cm_id->device->node_type != RDMA_NODE_IB_CA) 243 ret = -EADDRNOTAVAIL; 244 245 rdsdebug("addr %pI4 ret %d node type %d\n", 246 &addr, ret, 247 cm_id->device ? cm_id->device->node_type : -1); 248 249 rdma_destroy_id(cm_id); 250 251 return ret; 252 } 253 254 void rds_ib_exit(void) 255 { 256 rds_info_deregister_func(RDS_INFO_IB_CONNECTIONS, rds_ib_ic_info); 257 rds_ib_destroy_nodev_conns(); 258 ib_unregister_client(&rds_ib_client); 259 rds_ib_sysctl_exit(); 260 rds_ib_recv_exit(); 261 rds_trans_unregister(&rds_ib_transport); 262 } 263 264 struct rds_transport rds_ib_transport = { 265 .laddr_check = rds_ib_laddr_check, 266 .xmit_complete = rds_ib_xmit_complete, 267 .xmit = rds_ib_xmit, 268 .xmit_cong_map = NULL, 269 .xmit_rdma = rds_ib_xmit_rdma, 270 .xmit_atomic = rds_ib_xmit_atomic, 271 .recv = rds_ib_recv, 272 .conn_alloc = rds_ib_conn_alloc, 273 .conn_free = rds_ib_conn_free, 274 .conn_connect = rds_ib_conn_connect, 275 .conn_shutdown = rds_ib_conn_shutdown, 276 .inc_copy_to_user = rds_ib_inc_copy_to_user, 277 .inc_purge = rds_ib_inc_purge, 278 .inc_free = rds_ib_inc_free, 279 .cm_initiate_connect = rds_ib_cm_initiate_connect, 280 .cm_handle_connect = rds_ib_cm_handle_connect, 281 .cm_connect_complete = rds_ib_cm_connect_complete, 282 .stats_info_copy = rds_ib_stats_info_copy, 283 .exit = rds_ib_exit, 284 .get_mr = rds_ib_get_mr, 285 .sync_mr = rds_ib_sync_mr, 286 .free_mr = rds_ib_free_mr, 287 .flush_mrs = rds_ib_flush_mrs, 288 .t_owner = THIS_MODULE, 289 .t_name = "infiniband", 290 .t_type = RDS_TRANS_IB 291 }; 292 293 int __init rds_ib_init(void) 294 { 295 int ret; 296 297 INIT_LIST_HEAD(&rds_ib_devices); 298 299 ret = ib_register_client(&rds_ib_client); 300 if (ret) 301 goto out; 302 303 ret = rds_ib_sysctl_init(); 304 if (ret) 305 goto out_ibreg; 306 307 ret = rds_ib_recv_init(); 308 if (ret) 309 goto out_sysctl; 310 311 ret = rds_trans_register(&rds_ib_transport); 312 if (ret) 313 goto out_recv; 314 315 rds_info_register_func(RDS_INFO_IB_CONNECTIONS, rds_ib_ic_info); 316 317 goto out; 318 319 out_recv: 320 rds_ib_recv_exit(); 321 out_sysctl: 322 rds_ib_sysctl_exit(); 323 out_ibreg: 324 ib_unregister_client(&rds_ib_client); 325 out: 326 return ret; 327 } 328 329 MODULE_LICENSE("GPL"); 330 331